[iyonix-support] Re: Bugs in SCSIFS?

  • From: Dave Higton <davehigton@xxxxxxxxxxxxx>
  • To: iyonix-support@xxxxxxxxxxxxx
  • Date: Wed, 04 Jul 2007 22:47:35 +0100

In message <33caaef84e.martin@xxxxxxxxxxxxxxxxxxx>
          Martin Wuerthner <lists@xxxxxxxxxxxxxxx> wrote:

> In message <cce6a9f84e.davehigton@xxxxxxxxxxxxx>
>           Dave Higton <davehigton@xxxxxxxxxxxxx> wrote:
> 
> > [...] The widest range of USB devices includes devices with varying
> > numbers and types of partitions.
> 
> Of course, plus those with disk-level compression, with military 
> strength encryption plus steganography for plausible deniability. In 
> some labs you might also find devices using new formats under 
> development that have not been published yet. But what do the devices 
> look like that users come across 99% of the time? They have a single 
> FAT partition. Would it not make a huge amount of sense to start 
> supporting those and worry about all the other things later?

No.  If some fundamental part of RISC OS needs to be changed in order
to support any of this, then my vote is to change it in a way that
will permit all of the requirements that we can foresee.  It would
be foolish to change it once and then have to change it again later.
Let's do it once and get it right.

> > I understand the Filecore/FS/ImageFS structure.  I understand
> > that it needs an overhaul.
> 
> Sorry, I cannot understand your point. We have a primary problem that 
> needs to be solved, which is accessing >2GB FAT formatted USB storage 
> devices. This problem can be solved with a minor extension of the 
> current foreign disk format identification mechanism. There is 
> absolutely no need for revolutionary changes. As Druck wrote it is 
> straightforward for such a solution to support even multiple FAT 
> partitions without requiring any OS level partition support. Why are 
> you trying to make the issue more complicated than necessary?

I am, as always, searching for the simplest solution that will allow
all we want.  As with all jobs: if we do it right, we will only need
to do it once.

> As you wrote earlier, partitions are so simple. So, it will be trivial 
> to add support for them later. There is no need to worry about them 
> right now, apart from making sure that the code can start reading from 
> a given sector number.

Good, we're agreed.

Dave
---
To alter your preferences or leave the group, 
visit //www.freelists.org/list/iyonix-support
Other info via //www.freelists.org/webpage/iyonix-support

Other related posts: